Lens correction logic for image sensors

BACKGROUND OF THE INVENTION

1. Field of the Invention

Embodiments of the invention relate to optical devices, and more particularly to optical devices including digital imagers.

2. Description of the Related Art

Optical devices suffer from a position-dependent light intensity variation termed “vignetting.” Vignetting refers to an effect where an image is brighter towards the center of the image and fades towards the edges thereof. Vignetting is typically caused by characteristics of optical lenses of an optical device, which results in an uneven distribution of light intensity on the optical field (e.g., film or image sensor array) of the optical device.

Digital optical devices also suffer from vignetting. Digital optical devices typically include a lens system and an image sensor (or imager) that digitally captures images. Vignetting is particularly acute in low-cost digital optical devices that use an inexpensive optical lens, and have a relatively short distance between the lens and the image sensor thereof. When digital optical devices are in use, pixel data (e.g., luminance data) of a captured image can be processed to restore the brightness of the image, thereby compensating for vignetting in the digital optical devices. For example, pixel data can be multiplied by position-dependent gains to restore the brightness to the correct level. In some arrangements, look-up tables are employed to provide position-dependent gains across an array of pixels. However, such look-up tables need a large amount of memory, which increases the manufacturing costs.

To minimize memory requirements, pixel correction functions have been used for compensating for vignetting. Pixel correction functions typically provide position-dependent gains such that pixel data from pixels near edges of an image sensor array are multiplied by a larger correction gain than data from pixels near the center of the array. Various functions have been proposed as pixel correction functions, including polynomial functions, exponential functions, trigonometric functions, and hyperbolic cosine functions. Such pixel correction functions need to be calibrated for the position of the optical lens relative to the image sensor array in each individual digital optical device. The calibration process can be performed for individual digital optical devices during or after the manufacturing process. It would therefore be advantageous to provide a simplified calibration process for digital optical devices.

BRIEF DESCRIPTION OF THE DRAWINGS

The embodiments will be better understood from the Detailed Description of the Embodiments and from the appended drawings, which are meant to illustrate and not to limit the embodiments, and wherein:

<figref idrefs="DRAWINGS">FIG. 1</figref> is a flowchart illustrating a process of calibrating a pixel correction function for compensating for vignetting in a digital optical device in accordance with one embodiment;

<figref idrefs="DRAWINGS">FIG. 2</figref> is a schematic perspective view of a digital optical device set up for taking a reference image in accordance with one embodiment;

<figref idrefs="DRAWINGS">FIG. 3</figref> is a schematic top plan view of an optical field of a digital optical device with pixel calibration sample points in accordance with one embodiment;

<figref idrefs="DRAWINGS">FIG. 4A</figref> is a graph illustrating brightness variations due to vignetting along the horizontal direction of the optical field of <figref idrefs="DRAWINGS">FIG. 3</figref>;

<figref idrefs="DRAWINGS">FIG. 4B</figref> is a graph illustrating one embodiment of a pixel correction function with gains to compensate for the brightness variations of <figref idrefs="DRAWINGS">FIG. 4A</figref>;

<figref idrefs="DRAWINGS">FIG. 5A</figref> is a graph illustrating brightness variations due to vignetting along the vertical direction of the optical field of <figref idrefs="DRAWINGS">FIG. 3</figref>;

<figref idrefs="DRAWINGS">FIG. 5B</figref> is a graph illustrating one embodiment of a pixel correction function with gains to compensate for the brightness variations of <figref idrefs="DRAWINGS">FIG. 5A</figref>;

<figref idrefs="DRAWINGS">FIG. 6</figref> is a schematic block diagram of an imager employing a pixel correction function in accordance with one embodiment; and

<figref idrefs="DRAWINGS">FIG. 7</figref> is a schematic block diagram of a computer system having an imager employing a pixel correction function in accordance with one embodiment.

DETAILED DESCRIPTION OF EMBODIMENTS

In one embodiment, a method of calibrating a pixel correction function for compensating for vignetting in a digital optical device is provided. The digital optical device can be any optical device configured to digitally capture images, including, but not limited to, digital still cameras, digital video camcorders, film still cameras with digital image capturing features, and analog video camcorders with digital image capturing features. The optical device can include imagers of any suitable type, for example, CCD imagers or CMOS (semiconductor) imagers.

<figref idrefs="DRAWINGS">FIG. 1</figref> illustrates a process for calibrating a pixel correction function for an optical device according to one embodiment. An optical device can employ a pixel correction function including a first predetermined number of unknown constant values. In one embodiment, in order to determine the constant values in an anti-vignetting calibration process for the optical device, the optical device is exposed to a reference object in order to provide a reference image. A second predetermined number of sample points are selected from the reference image for determining the unknown constant values. In one embodiment, the second predetermined number is greater than the first predetermined number by one (1). In another embodiment, the second predetermined number is equal to the first predetermined number. This method simplifies the calibration process, and thus reduces the manufacturing costs.

In one embodiment, an imager uses a function of the product of two hyperbolic cosine functions as a pixel correction function. The pixel correction function serves to provide two-dimensional position-dependent gains for vignetting correction. The pixel correction function is configured to provide gains that approximate correction across the optical field of an imager.

The pixel correction function can include a horizontal component g(x) for gains along the horizontal (x) direction of the optical field. The horizontal component g(x) can be represented by Equation 1. <br /><i>g</i>(<i>x</i>)=cos <i>h</i>(<i>s</i>_x(<i>x−c</i>_x))  (1)<br /> In Equation 1, s_xis a constant horizontal scaling factor in the x direction and c_xis a constant horizontal center value in the x direction.

The pixel correction function can also include a vertical component g(y) for gains along the vertical (y) direction of the optical field. The vertical component g(y) can be represented by Equation 2. <br /><i>g</i>(<i>y</i>)=cos <i>h</i>(<i>s</i>_y(<i>y−c</i>_y))  (2)<br /> In Equation 2, s_yis a constant vertical scaling factor in the y direction and c_yis a constant vertical center value in the y direction.

The pixel correction function is the product of the horizontal component g(x) of Equation 1 and the vertical component g(y) of Equation 2, which provides two-dimensional gains for vignetting correction. The pixel correction function is represented by Equation 3. <br /><i>g</i>(<i>x,y</i>)=<i>g</i>(<i>x</i>)<i>g</i>(<i>y</i>)=cos <i>h</i>(<i>s</i>_x(<i>x−c</i>_x))cos <i>h</i>(<i>s</i>_y(<i>y−c</i>_y))  (3)<br /> The constant values s_x, s_y, c_x, and c_yare unknown and can be determined by a calibration process so that the function can correct for vignetting for the configurations of the lens and imager sensors of a digital optical device.

In one embodiment, the constant values of Equation 3 can be determined by the observation of a predetermined number of selected sample points. Each of the horizontal or vertical components g(x), g(y) of Equation 3 has two unknowns and can be fixed by the observation of three sample points in comparison to known intensity of the image at the sample points. In one embodiment, the constant values can be determined by sampling three points in each dimension, and finding a unique solution of the equation which satisfies the observations. The solution can be arrived at by a closed-form solution (by which at least one solution can be expressed analytically in terms of a bounded number of certain “known” functions) or by an iterative method.

By using one of the sample points for both of the horizontal and vertical components g(x), g(y) of Equation 3, the calibration can be performed with only 5 sample points, while still obtaining a minimum of 3 sample points along each of two intersecting lines. In one embodiment, the sample points can be selected from two intersecting lines. The intersecting lines can be orthogonal and can extend in the horizontal and vertical direction, respectively. At least two of the sample points can have the same value in the vertical or horizontal dimension.

Referring to <figref idrefs="DRAWINGS">FIG. 1</figref>, a process for calibrating a pixel correction function for an optical device is described below according to one embodiment. The calibration process can be performed during or after the manufacturing process of the optical device as long as the optical device has been equipped with its essential components, i.e., optical lens and image sensors. In the illustrated embodiment, the calibration process can start with providing a reference image (block <b>110</b>). In one embodiment, the reference image can be provided by capturing an image of a reference object. The capturing of a reference object can be accomplished in numerous fashions depending upon the desired use of the imager.

In one embodiment, referring to <figref idrefs="DRAWINGS">FIG. 2</figref>, a reference object <b>250</b> is a flat object illuminated with a stationary uniform white light source. In certain embodiments, this step can also be performed in a dark room to avoid ambient light interference. The reference object <b>250</b> is aligned to face a digital optical device <b>200</b> that includes an optical lens system <b>210</b> and an example of an imager <b>220</b>. The optical lens system <b>210</b> can include one lens or a stack of lenses. The imager <b>220</b> is on a die and includes an array of image sensors <b>221</b> in a matrix form in the illustrated embodiment. It should be noted, however, that although the imager <b>220</b> has been shown as an array of image sensors <b>221</b>, the imager <b>220</b> can be represented in numerous other arrangements as well. A reference image of the reference object <b>250</b> is captured by the imager <b>220</b>.

In another embodiment, the calibration process can use a single light source that is moved relative to the imager <b>220</b> during the calibration process to illuminate different sample points. In this embodiment, the light source is directed to illuminate one sample point at a time, and a reference image is captured for the one sample point. Then, the light source can be moved or directed to illuminate another sample point with substantially the same intensity of light, and then another reference image is captured for the other sample point. This step is repeated until reference images for all the samples points are captured. Then, the reference images can be processed to provide the brightness levels of the sample points. In some embodiments, the reference images can be consolidated into a single reference image similar to the one captured with the uniform white light source described above. This process does not depend on providing a uniform flat image as big as the imager; rather a single light emitting diode (LED) can provide the known brightness identically at each sample point across the imager.

In other embodiments, the calibration process can use static multiple light sources with substantially the same light intensity for illuminating the sample points. For example, five LEDs can be used to illuminate five sample points with substantially the same intensity of light. In one embodiment, a single reference image can be captured with all the LEDs on at the same time. In another embodiment, five separate reference images can be captured with one of the LEDs on at a time, and then the reference images can be processed to provide the brightness levels of the sample points. The reference images taken in this way may be merged into a single reference image similar to the one captured with the uniform white light source described above. As noted, with respect to the movable LED embodiment, it is easier to provide five small identical light sources than one big uniform light source. A skilled artisan will appreciate that various other methods can be used to provide reference images.

Referring again to <figref idrefs="DRAWINGS">FIG. 1</figref>, at block <b>120</b>, the brightness (or intensity) levels of samples points are obtained from the reference image(s). In one embodiment, the total number SP of the sample points is represented by Equation 4. <br /><i>SP=N+</i>1  (4)

In Equation 4, N is the total number of constant values in a pixel correction function used for the imager. In the illustrated embodiment in which the pixel correction function is the product of two hyperbolic cosine functions with a total of four constant values, N is 4, and thus the total number SP of the sample points is 5. Whether employing a single reference image covering the field, or multiple reference images for individual sample points, because a small number of sample points are used for the calibration, the method is less reliant on a uniformly bright, “flat” image covering the whole field.

In one embodiment, the sample points are selected such that at least two of the sample points have the same horizontal or vertical value. Referring to <figref idrefs="DRAWINGS">FIG. 3</figref>, five sample points A<b>1</b>, A<b>2</b>, A<b>3</b>, B<b>1</b>, and B<b>2</b> are selected for calibrating the pixel correction function of Equation 3. The five sample points A<b>1</b>, A<b>2</b>, A<b>3</b>, B<b>1</b>, and B<b>2</b> have the following horizontal and vertical values on the reference image.

A<b>1</b>: x₁, y₂

A<b>2</b>: x₂, y₂

A<b>3</b>: x₃, y₂

B<b>1</b>: x₂, y₁

B<b>2</b>: x₂, y₃

Among the five sample points, a horizontal group or line of three sample points A<b>1</b>, A<b>2</b>, A<b>3</b> have the same vertical (y) value y₂. A vertical group or line of three sample points B<b>1</b>, A<b>2</b>, B<b>2</b> have the same horizontal (x) value x₂.

The brightness levels of the horizontal group of sample points A<b>1</b>, A<b>2</b>, A<b>3</b> can be obtained from the reference image(s) as shown in <figref idrefs="DRAWINGS">FIG. 4A</figref>. As shown in <figref idrefs="DRAWINGS">FIG. 4A</figref>, the sample point A<b>2</b> near the center of the reference image is higher in brightness than the other sample points A<b>1</b>, A<b>2</b> near the two side edges of the image. The brightness levels of the vertical group of sample points B<b>1</b>, A<b>2</b>, B<b>2</b> can be obtained as shown in <figref idrefs="DRAWINGS">FIG. 5A</figref>. As shown in <figref idrefs="DRAWINGS">FIG. 5A</figref>, the sample point A<b>2</b> near the center of the reference image is higher in brightness than the other sample points B<b>1</b>, B<b>2</b> near the upper and lower edges of the image.

After obtaining the brightness levels of the sample points, the reciprocal values of the brightness levels are obtained at block <b>130</b>. <figref idrefs="DRAWINGS">FIG. 4B</figref> illustrates correction gains which are the reciprocal values <b>41</b>, <b>42</b>, <b>43</b> of the brightness levels of the horizontal group of sample points A<b>1</b>, A<b>2</b>, A<b>3</b>. <figref idrefs="DRAWINGS">FIG. 5B</figref> illustrates correction gains which are the reciprocal values <b>51</b>, <b>52</b>, <b>53</b> of the brightness levels of the vertical group of sample points B<b>1</b>, A<b>2</b>, B<b>2</b>.

At block <b>140</b>, using the reciprocal values <b>41</b>, <b>42</b>, <b>43</b> of the brightness levels of the horizontal group, the constant values s_x, c_xof the horizontal component g(x) of Equation 3 can be obtained by a closed-form solution or by an iterative method (e.g., Levenberg-Marquardt method). Likewise, using the reciprocal values <b>51</b>, <b>52</b>, <b>53</b> of the brightness levels of the vertical group, the constant values s_y, c_yof the vertical component g(y) of Equation 3 can be obtained by a closed-form solution or by an iterative method (e.g., Levenberg-Marquardt method). In the illustrated embodiment, the sample point A<b>2</b> is in both the horizontal and vertical groups, which allows minimizing the total number of sample points for determining the constant values of the pixel correction function of Equation 3.

In another embodiment, an imager can use a pixel correction function with an overall gain factor. An overall gain factor serves to provide the absolute value of a gain applied to each pixel. On the other hand, a pixel correction function without an overall gain factor (e.g., Equation 3) may only provide normalized gains for vignetting correction. Normalized gains provide relative adjustments of values between pixels. Such a pixel correction function can be a hyperbolic cosine function represented by Equation 5. <br /><i>g</i>(<i>x,y</i>)=<i>A*g</i>(<i>x</i>)<i>g</i>(<i>y</i>)=<i>A</i>*cos <i>h</i>(<i>s</i>_x(<i>x−c</i>_x))cos <i>h</i>(<i>s</i>_y(<i>y−c</i>_y))  (5)

In Equation 5, A is an overall gain factor, and is a constant value which can be determined by a calibration process described above in connection with <figref idrefs="DRAWINGS">FIG. 1</figref>. Other constant values s_x, s_y, c_x,and c_yare as described above with respect to Equations 1-3. The function of Equation 5 has a total of five constant values.

In yet another embodiment, a pixel correction function with an overall gain factor can be a polynomial function represented by Equation 6.

<maths id="MATH-US-00001" num="00001"><math overflow="scroll"><mtable><mtr><mtd><mrow><mrow><mi>f</mi><mo>⁡</mo><mrow><mo>(</mo><mrow><mi>x</mi><mo>,</mo><mi>y</mi></mrow><mo>)</mo></mrow></mrow><mo>=</mo><mrow><munderover><mo>∑</mo><mrow><mi>i</mi><mo>=</mo><mn>0</mn></mrow><mn>4</mn></munderover><mo>⁢</mo><mrow><munderover><mo>∑</mo><mrow><mi>j</mi><mo>=</mo><mn>0</mn></mrow><mn>4</mn></munderover><mo>⁢</mo><mrow><msub><mi>k</mi><mi>ij</mi></msub><mo>⁢</mo><msup><mi>x</mi><mi>i</mi></msup><mo>⁢</mo><msup><mi>y</mi><mi>j</mi></msup></mrow></mrow></mrow></mrow></mtd><mtd><mrow><mo>(</mo><mn>6</mn><mo>)</mo></mrow></mtd></mtr></mtable></math></maths>

In Equation 6, f(x, y) represents a brightness gain to be multiplied by the pixel data from a position (x, y) on a pixel array. The variable k_ijrepresents the coefficients of the brightness gain; i and j are indices to k (the matrix of coefficients); (x, y) are the coordinates of the pixels.

Equation 6 has 25 unknown coefficients (or constants) because the polynomial function is in the order of 5 in both x and y directions. Equation 6 can be solved by a system of 25 linear equations using 25 sample points.

In the embodiments in which a pixel correction function includes an overall gain factor, the total number SP of the sample points for a calibration process is represented by Equation 7. <br />SP=N  (7)

In Equation 7, N is the total number of constant values in a pixel correction function with an overall gain factor. The total number SP of the sample points for the calibration process is equal to the total number of constant values in a pixel correction function used for the imager. For example, with respect to Equation 5, the total number of constant values in the pixel correction function is 5, and therefore, the total number SP of the sample points for the calibration process is also 5. With respect to Equation 6, the total number of constant values in the pixel correction function is 25, and therefore, the total number SP of the sample points for the calibration process is 25.

In the embodiments described above, only three pixel correction functions are described as examples. A skilled artisan will, however, appreciate that various other pixel correction functions can be calibrated using the calibration method described above with reference to <figref idrefs="DRAWINGS">FIG. 1</figref>.

The constant values obtained from the process described above can be stored in a memory (e.g., ROM) of the imager. The pixel correction function with the determined constant values is used in compensating for vignetting during operation of the imager.

<figref idrefs="DRAWINGS">FIG. 6</figref> illustrates a block diagram of an imager <b>600</b> constructed in accordance with one of the embodiments described above. The illustrated imager <b>600</b> is a CMOS or semiconductor imager. The imager <b>600</b> contains an array of pixels <b>620</b> and employs a calibration method for providing a pixel correction function according to the embodiment shown in <figref idrefs="DRAWINGS">FIG. 1</figref>. Attached to the pixel array <b>620</b> is signal processing circuitry for controlling the pixel array <b>620</b>. The pixel cells of each row in the array <b>620</b> are all turned on at the same time by a row select line, and the pixel cells of each column are selectively output by respective column select lines. A plurality of row select and column select lines are provided for the entire array <b>620</b>. The row lines are selectively activated by a row driver <b>645</b> in response to a row address decoder <b>655</b>. The column select lines are selectively activated by a column driver <b>660</b> in response to a column address decoder <b>670</b>. Thus, a row and column address is provided for each pixel cell.

The imager <b>600</b> is operated by a timing and control circuit <b>652</b>, which controls the address decoders <b>655</b>, <b>670</b> for selecting the appropriate row and column lines for pixel readout. The control circuit <b>652</b> also controls the row and column driver circuitry <b>645</b>, <b>660</b> such that they apply driving voltages to the drive transistors of the selected row and column lines. The pixel column signals, which typically include a pixel reset signal V_rstand a pixel image signal V_sig, are output to the column driver <b>660</b>, on output lines, and are read by a sample and hold circuit <b>661</b>. V_rstis read from a pixel cell immediately after the pixel cell's floating diffusion region is reset. V_sigrepresents the amount of charge generated by the photosensitive element of the pixel cell in response to applied light during an integration period. A differential signal (V_rst−V_sig) is produced by a differential amplifier <b>662</b> for each readout pixel cell. The differential signal is digitized by an analog-to-digital converter <b>675</b> (ADC). The analog-to-digital converter <b>675</b> supplies the digitized pixel signals to an image processor <b>680</b>, which forms and outputs a digital image. In one embodiment, the image processor <b>680</b> can use a pixel correction function calibrated with the method described above with reference to <figref idrefs="DRAWINGS">FIG. 1</figref>. The pixel correction function can be stored in a memory which is associated with or part of the image processor <b>680</b>.

<figref idrefs="DRAWINGS">FIG. 7</figref> illustrates a processor system <b>700</b> that includes an imager <b>600</b> constructed in accordance with an embodiment. As discussed above, the imager <b>600</b> contains an array of pixels and employs a method of calibrating a pixel correction function according to the embodiment shown in <figref idrefs="DRAWINGS">FIG. 1</figref>. The system <b>700</b> has digital circuits that could include image sensor devices. Without being limiting, such a system could include a computer system, camera system, scanner, machine vision, vehicle navigation, video phone, surveillance system, auto focus system, star tracker system, motion detection system, image stabilization system, and other image sensing and/or processing system.

The system <b>700</b>, for example a camera system, generally includes a central processing unit (CPU) <b>702</b>, such as a microprocessor, that communicates with an input/output (I/O) device <b>706</b> over a bus <b>704</b>. The imager <b>600</b> also communicates with the CPU <b>702</b> over the bus <b>704</b>. The processor system <b>700</b> also includes a random access memory (RAM) <b>710</b>, and may include a removable memory <b>715</b>, such as a flash memory, which also communicates with the CPU <b>702</b> over the bus <b>704</b>. The imager <b>600</b> may be combined with a processor, such as a CPU, digital signal processor, or microprocessor, with or without memory storage on a single integrated circuit or on a different chip than the processor. In certain embodiments, the CPU <b>702</b> in conjunction with the RAM <b>710</b> or the removable memory <b>715</b> can process pixel data for vignetting correction. A skilled artisan will appreciate that it is also possible to process pixel data for vignetting correction on a separate computing device or system. The CPU <b>702</b> can use a pixel correction function calibrated with the method described above with reference to <figref idrefs="DRAWINGS">FIG. 1</figref>.

In the embodiments described above, a pixel correction function can be calibrated with a small number of sample points. This configuration allows simplifying calibration processes for individual optical devices, and thus reduces the manufacturing costs.

One embodiment is a method of calibrating a pixel correction function for compensating for vignetting in an optical device. The method includes exposing an optical device to at least one reference object so as to generate at least one reference image of the at least one reference object. The method also includes providing a pixel correction function including a first number of unknown constant values. The method further includes providing pixel data of a second number of sample points on the at least one reference image. The second number is equal to the first number or the first number plus one. The method also includes determining the constant values using the pixel data of the second number of sample points.

Another embodiment is a method of calibrating a pixel correction function for compensating for vignetting in an optical device. The method includes exposing an optical device to a reference object. The optical device includes an imager including an array of pixels. The method also includes capturing at least one image of the reference object so as to generate pixel data of the at least one image; providing pixel data of sample points taken along two intersecting lines of the at least one image; and calibrating a pixel correction function with the pixel data of the sample points.

Yet another embodiment is a method of calibrating a pixel correction function for vignetting compensation in an optical device. The method includes providing at least one reference object including a selected number of portions having substantially the same brightness simultaneously or sequentially. The method also includes capturing at least one image of the at least one reference object with an optical device so that the at least one image includes the portions of the reference object. The optical device is configured to use a pixel correction function for vignetting correction. The method further includes providing pixel data of sample points on the at least one reference image. Each of the sample points corresponds to a respective one of the portions of the reference object. The method includes calibrating the pixel correction function with the pixel data.

Another embodiment is an optical device including a lens system; an image sensor array configured to receive light passing through the lens; and a processor configured to compensate for vignetting associated with at least one of the lens and the image sensor array. The processor is configured to use a pixel correction function including a first number of constant values. The pixel correction function is calibrated by a method including: exposing the optical device to a reference object; capturing at least one image of the reference object with the image sensor array so as to generate pixel data of the image; providing pixel data of sample points taken along two intersecting lines of the image; and calibrating the pixel correction function with the pixel data of the sample points.

Although this invention has been described in terms of certain embodiments, other embodiments that are apparent to those of ordinary skill in the art, including embodiments that do not provide all of the features and advantages set forth herein, are also within the scope of this invention. Moreover, the various embodiments described above can be combined to provide further embodiments. In addition, certain features shown in the context of one embodiment may be incorporated into other embodiments as well. Accordingly, the scope of the present invention is defined only by reference to the appended claims.